No abstract

No abstract

Abstract. Orchids of the genus Ophrys (Orchidaceae) are pollinated by male bees and wasps through sexual deception. The Ophrys sphegodes group encompasses several closely related species that differ slightly in floral morphology and are pollinated by different solitary bee species. Populations representing different species of the O. sphegodes group often flower simultaneously in sympatry. To test whether gene flow across the species boundaries occurs in these sympatric populations, or whether they are reproductively isolated, we examined the distribution of genetic variation within and among populations and species of this group. We collected at each of five different localities in southern France and Italy two sympatric, co-flowering Ophrys populations, representing six Ophrys species in total. The six microsatellite loci surveyed were highly variable. Genetic differentiation among geographically distant populations of the same species was lower than differentiation among sympatric populations of different species. However, the strength of genetic differentiation among species was among the lowest reported for orchids. Genotype assignment tests and marker-based estimates of gene flow revealed that gene flow across species boundaries occurred and may account for the low observed differentiation among species. These results suggest that sexual deceit pollination in Ophrys may be less specific than thought, or that rare mistakes occur.

Orchids of the genus Ophrys (Orchidaceae) are pollinated by male bees and wasps through sexual deception. The Ophrys sphegodes group encompasses several closely related species that differ slightly in floral morphology and are pollinated by different solitary bee species. Populations representing different species of the O. sphegodes group often flower simultaneously in sympatry. To test whether gene flow across the species boundaries occurs in these sympatric populations, or whether they are reproductively isolated, we examined the distribution of genetic variation within and among populations and species of this group. We collected at each of five different localities in southern France and Italy two sympatric, co-flowering Ophrys populations, representing six Ophrys species in total. The six microsatellite loci surveyed were highly variable. Genetic differentiation among geographically distant populations of the same species was lower than differentiation among sympatric populations of different species. However, the strength of genetic differentiation among species was among the lowest reported for orchids. Genotype assignment tests and marker-based estimates of gene flow revealed that gene flow across species boundaries occurred and may account for the low observed differentiation among species. These results suggest that sexual deceit pollination in Ophrys may be less specific than thought, or that rare mistakes occur.

Gymnadenia conopsea s.l. is a common orchid in central Europe, where early- and late-flowering populations can be distinguished. The early-flowering form is recognized as subspecies conopsea and the late-flowering form as subspecies densiflora. The two subspecies can occur in sympatry, but their flowering periods are separated. We investigated whether early- and late-flowering subspecies are genetically differentiated, whether they diverged once or repeatedly, and we tried to identify potential evolutionary forces involved in the divergence of the two subspecies. We used genetic markers to estimate genetic divergence within and among populations of early- and late-flowering G. conopsea, and to reconstruct their evolutionary history. In addition, we assessed morphological variation between subspecies. Allozyme variation indicated that subspecies conopsea was significantly more variable than ssp. densiflora and that gene flow among populations of ssp. conopsea was higher than among populations of ssp. densiflora. Gene flow between subspecies was low, indicating that the difference in flowering phenology represented an effective barrier to gene flow. A neighbor-joining tree based on allozyme frequencies indicated that early- and late- flowering populations did not diverge repeatedly in sympatry. Levels of cpDNA variation were generally low, even between G. conopsea s.l. and Gymnadenia odoratissima, chosen as an outgroup. Four cpDNA haplotypes were found, which differed only in the number of microsatellite repeats. Their distribution among subspecies of G. conopsea s.l. and G. odoratissima indicates that microsatellite haplotypes have evolved repeatedly, and their occurrence in different taxa thus represents a homoplasy. Floral characters were variable within and among populations and subspecies but did not consistently separate early- from late-flowering populations. A weak separation between subspecies was found in vegetative characters that presumably reflected habitat and competitive differences experienced by early- and late-flowering populations.